Abortion

States first began to outlaw abortion around 1820, and by 1880, most states across the United States had banned the procedure, with some exceptions if the life of the mother was at risk (Petchesky Reference Petchesky1984; Reagan Reference Reagan1997; Haugeberg Reference Haugeberg2017). During this time period, opposition to abortion emphasized traditional gender norms, and also had a racialized component, with some anti-abortion advocates concerned about falling Protestant birth rates amid a rising immigrant population (Reagan Reference Reagan1997).

In 1973, the Supreme Court's decision in the Roe v. Wade case legalized the procedure, transforming the politics of the abortion debate. After Roe, the anti-abortion movement switched its focus away from preserving the traditional family unit toward an emphasis on the status of embryos and fetuses (Dolgin, Fisher and Shapiro Reference Dolgin, Fisher and Shapiro2008). The contemporary moral debate over abortion, including discussions about “fetal personhood” and whether an embryo deserves the same protections as other human life, is relatively recent; while today's opponents often portray abortion as a crime, even invoking the term “infanticide,” this rhetoric did not characterize earlier abortion debates (Tribe Reference Tribe1992; Ferree Reference Ferree2002).

Contemporary pro-life advocates are more likely to view embryos at the point of conception as human life, while pro-choice advocates typically take a gradualist approach to personhood, viewing the embryo as potential human life that develops into personhood at some point after conception (Luker Reference Luker1984). These underlying attitudes influence opinions about whether it is morally appropriate to destroy or discard human embryos. Individuals who view embryos as morally equivalent to human beings argue that these embryos should have the same rights as people, meaning that it is immoral to destroy or dispose of them. On the other extreme, some pro-choice advocates see embryos as having no moral status whatsoever and believe they should not warrant any special protection (Hudson, Scott and Faden Reference Hudson, Scott and Faden2005). Consistent with this approach, some feminists argue that the idea of fetal personhood is objectionable “because what makes human life distinct is its capacity for consciousness and sociability,” whereas fetal personhood “demean(s) human life and the moral value of consciousness” (Petchesky Reference Petchesky1984, 341). Finally, others take a middle view, that human embryos deserve some moral treatment, but to a lesser extent than a child outside of the womb. This perspective suggests that under certain conditions, destroying embryos may be morally acceptable.

In contrast to IVF, which is commonly seen as a procedure reserved for upper class white women (Bell Reference Bell2009), abortion rates are higher among unmarried, low-income, and African-American women (Finer and Henshaw Reference Finer and Henshaw2006). Thus, in addition to being a morality politics issue, there are also distinct intersectional implications to the abortion debate compared with IVF and ESC research.

Generally speaking, existing scholarship suggests that religious views and practice are more influential than demographics in shaping abortion attitudes. A strong and persistent relationship exists between religiosity and abortion attitudes, with those who are more devout more likely to be pro-life (Luker Reference Luker1984; Wilcox and Norrander Reference Wilcox, Norrander, Norrander and Wilcox2002; Jelen and Wilcox Reference Jelen and Wilcox2003; Jelen Reference Jelen2014; Lizotte Reference Lizotte2015). Religious denomination also plays a role in shaping public opinion on this issue, as do partisanship and ideology (Jelen and Wilcox Reference Jelen and Wilcox2003; Cook, Jelen and Wilcox Reference Cook, Jelen and Wilcox1992). Most research suggests that there is no relationship between gender and abortion attitudes, and some scholars find that men may even hold more liberal views (Cook, Jelen and Wilcox Reference Cook, Jelen and Wilcox1992; Huddy, Cassese and Lizotte Reference Huddy, Cassese, Lizotte, Wolbrecht, Beckwith and Baldez2008; Jelen and Wilcox Reference Jelen and Wilcox2003; Wilcox and Norrander Reference Wilcox, Norrander, Norrander and Wilcox2002; Jelen Reference Jelen2014), with women more likely to view abortion as morally wrong (Scott Reference Scott1989). However, Scott and Schuman (Reference Scott and Schuman1988) find that among pro-choice supporters specifically, the issue of abortion rights is of greater importance to women than to men, while Lizotte (Reference Lizotte2015) argues that religiosity accounts for the unexpected null relationship between gender and abortion attitudes. Although African Americans used to be more likely to oppose abortion, this association appears to have diminished over time, and possibly even reversed (Strickler and Danigelis Reference Strickler and Danigelis2002; Jelen and Wilcox Reference Jelen and Wilcox2003; Carter, Carter and Dodge Reference Carter, Carter and Dodge2009). I seek to better understand the extent to which these relationships hold across a range of embryonic politics issues. Given the strong relationship between religion and morality, I am especially interested in the extent to which religion similarly influences attitudes about other embryonic politics issues.

ESC Research

Stem cell research involves the use of certain types of cells to research and develop cures for various diseases. Because stem cells have regenerative capabilities, and the unique ability to transform into any of the types of cells in the human body (“pluripotent”), these cells provide possibilities for cell-based therapies and treatments for a wide range of diseases, including spinal cord injuries, diabetes, cardiovascular disease, and beyond (NIH Stem Cell Information Home Page 2016).

Medical research is conducted on two categories of stem cells: non-ESCs, which are found in various organs and tissues in the human body; and ESCs, which are derived from embryos fertilized in vitro that are donated through an informed consent process (NIH Stem Cell Information Home Page 2016). Given the focus of this paper, I look specifically at attitudes toward ESC research.Footnote ²

Both the news media and political elites portray ESC research as a moral issue (Nisbet, Brossard and Kroepsch Reference Nisbet, Brossard and Kroepsch2003; Ho, Brossard and Scheufele Reference Ho, Brossard and Scheufele2008; Clifford and Jerit Reference Clifford and Jerit2013) because, like abortion, this technique raises questions about the status of embryos.Footnote ³ ESC research has also been described as a “wedge” issue, with the potential to pull voters from one party to another (Hillygus and Shields Reference Hillygus and Shields2014). Among abortion opponents who believe that embryos are the moral equivalent of human beings from conception, we would expect to see similar opposition to ESC research. However, while the goal of abortion is arguably the destruction of life, the goal of ESC research is to create new cures that will lessen human suffering (Meilander Reference Meilander2001). This distinction shifts the ethical debate to questions about whether it is appropriate to sacrifice embryos for the greater good of the community (McGee and Caplan Reference McGee and Caplan1999).

Disputes over ESC research reflect partisan disagreements over abortion, as well as broader debates about the role of government, religion, and science in society (Thompson Reference Thompson2013). Much of the policy debate over ESC research has focused primarily on government funding, including whether the government should support a procedure to which some citizens are morally opposed. For instance, in 2001, Republican President George W. Bush limited federal funding for ESC research to existing stem cell lines, banning research on new cell lines or newly derived cells. Democratic President Barack Obama overturned this restriction in 2009.

As with the abortion debate, the Catholic Church and some other Christian groups actively oppose ESC research, arguing that destroying an embryo would be “essentially equivalent to murder” (Nisbet, Brossard and Kroepsch Reference Nisbet, Brossard and Kroepsch2003, 135). At the same time, other religious groups have declined to take a position on the issue, and some Jewish and more liberal Christian groups have advocated in support of ESC research (Vestal Reference Vestal2008).

With respect to public opinion, while some similarities are evident in attitudes about these procedures, not all abortion opponents consistently oppose ESC research (Dolgin, Fisher and Shapiro Reference Dolgin, Fisher and Shapiro2008). Indeed, even at the elite level, some pro-life politicians have publicly supported ESC research. For instance, in 2005, then-Senate Majority Leader Bill Frist (R-TN) publicly broke with President Bush, citing the “truly magnificent, truly remarkable properties” of stem cells (Connolly Reference Connolly2005). Similarly, I anticipate that overall, a smaller share of the population will view ESC research, as compared with abortion, as morally objectionable and a greater share will view it as not a moral issue. I also expect that, compared with abortion attitudes, there will be a weaker relationship between attitudes about ESC research and religiosity or religious denomination.

In Vitro Fertilization

Individuals in the United States increasingly use assisted reproductive technologies (ARTs) to conceive. In 1996, the first year in which the Centers for Disease Control began to collect comprehensive data, 64,036 ART cycles were performed, and 20,597 babies were born from these procedures (Sunderam et al. Reference Sunderam, Kissin, Crawford, Folger, Jamieson, Warner and Barfield2015). These figures have increased more than threefold in the last 20 years. Yet, even as the use of ARTs has grown substantially, our understanding of attitudes toward these procedures remains limited.

Similar to other policies involving human embryos, questions about the morality of IVF likewise focus on when human life begins, and whether the destruction of an embryo is equivalent to the destruction of human life. In a typical IVF cycle, more embryos are produced than are actually used, leading to a debate over what to do with surplus fertilized embryos. Options include freezing them indefinitely, donating them for ESC research, making them available for adoption, or disposal (Nachtigall et al. Reference Nachtigall, Becker, Friese, Butler and MacDougall2005). Other questions relate to the number of embryos transferred during IVF: while the American Society of Reproductive Medicine (ASRM) has clear guidelines, no federal laws exist to prevent large numbers of embryos from being implanted.Footnote ⁴ The transfer of multiple embryos can result in multifetal pregnancies, leading to debates over techniques by which one or more fetuses are “reduced” or aborted.Footnote ⁵

Class inequality significantly limits access to IVF, which is not covered by most private insurance plans or Medicaid. As of March 2017, the ASRM (ND) listed the cost of a single cycle of IVF as $12,400, while another study placed the median cost at $19,200 (Wu et al. Reference Wu, Odisho, Washington, Katz and Smith2014). Despite the fact that racial and ethnic minorities are disproportionately affected by infertility, other structural and cultural factors also result in inequalities of access (Spar Reference Spar2006; Inhorn and Birenbaum-Carmeli Reference Inhorn and Birenbaum-Carmeli2008; Culley, Hudson and van Rooj Reference Culley, Hudson and van Rooj2013). Another indicator of the complex politics behind ARTs is the mixed reaction of feminist groups (Donchin Reference Donchin1986; Franklin Reference Franklin, Ginsburg and Rapp1995). While most feminist organizations have been strong advocates for reproductive rights, they have largely avoided embracing the cause of expanding access to IVF (Donchin Reference Donchin1996; Harwood Reference Harwood2007; Heidt-Forsythe Reference Heidt-Forsythe2013; Mottier Reference Mottier, Waylen, Celis, Kantola and Weldon2013).

When IVF was still in its early stages, the National Institute of Health created an Ethics Advisory Board (EAB) to evaluate this procedure. The Board's 1979 report concluded that the “human embryo is entitled to profound respect; but this respect does not necessarily encompass the full legal and moral rights attributed to persons” (quoted in Bonnicksen Reference Bonnicksen1989, 79). Following an outpouring of public feedback against these conclusions, IVF became “a pariah attached to the politically dangerous issue of abortion” (Bonnicksen Reference Bonnicksen1989, 81). Since this report, the federal government has largely avoided involvement in debates surrounding ARTs, in contrast to other policies related to embryos (Bonnicksen Reference Bonnicksen1989, Jasanoff Reference Jasanoff2011, Stapleton and Skinner Reference Stapleton and Skinner2015). Given the exponential growth of IVF since 1979, it remains unclear whether the early link identified between IVF and abortion persists, and to my knowledge, no empirical research has examined this question.

In general, questions of embryo disposal and destruction in IVF have not been subject to the same fierce political debates surrounding abortion and, to a lesser extent, ESC research. Rather, ARTs occupy an intermediate policy space, “stretch(ing) the boundaries of current conceptualizations of morality policy,” while also incorporating business and scientific interests (Heidt-Forsythe Reference Heidt-Forsythe2013, 20).

Scholarship on public opinion toward abortion is expansive, and a substantial body of research also exists on attitudes regarding stem cell research. Far less research examines public opinion about IVF. Adashi et al. (Reference Adashi, Cohen, Hamberger, Jones, De Kretser, Lunenfeld, Rosenwaks and Van Steirteghem2000) report on the first large-scale international survey on infertility, but their primary emphasis is cross-national. Other research on ART attitudes in the United States is based on limited samples, but points to a relationship between partisanship/ideology and IVF, with Democrats more likely to support the procedure and conservatives more likely to oppose (Shreffler, Johnson and Scheublec Reference Shreffler, Johnson and Scheublec2010; Sigillo, Miller and Weiser Reference Sigillo, Miller and Weiser2012; Fortin and Abele Reference Fortin and Abele2016). Fortin and Abele (Reference Fortin and Abele2016) find that religiosity, measured by attendance at religious services, is associated with more negative attitudes about IVF, but that Catholic women are actually more supportive compared with non-Catholics, a counterintuitive finding given the Church's strong opposition to abortion.

Existing research on the range of embryonic politics issues yields four hypotheses.

H1: Given the hybrid position of ARTs, their distinct history, and rapid growth, I anticipate that compared with abortion and ESC research, far fewer people will describe IVF as morally wrong.

At the same time, I expect that some respondents will hold constrained, or consistent, attitudes about embryonic politics. In particular, religion strongly influences attitudes about the status of embryos, particularly among conservative Christians (Luker Reference Luker1984; Wilcox and Norrander Reference Wilcox, Norrander, Norrander and Wilcox2002; Jelen and Wilcox Reference Jelen and Wilcox2003; Ho, Brossard and Scheufele Reference Ho, Brossard and Scheufele2008; Liu and Priest Reference Liu and Priest2009; Fortin and Abele Reference Fortin and Abele2016). There is also a relationship between religion and “constrained political attitudes” (Pearson-Merkowitz and Gimpel Reference Pearson-Merkowitz, Gimpel, Guth, Kellstedt and Smidt2009). This research yields two additional hypotheses.

H2: There will be a positive relationship between religiosity, or strength of religious feelings, and embryonic politics attitudes. More devout respondents will be more likely to describe all procedures as morally wrong and will be more likely to oppose abortion, followed by ESC research and then IVF.

Both Catholics and Evangelical Protestants strongly emphasize that life begins at the moment of conception (Ellison, Echevarría and Smith Reference Ellison, Echevarría and Smith2005). At the elite level, office holders who are Evangelical Christian and those representing Evangelical districts are most conservative on culture wars issues, with Catholics and mainline Protestants in the middle, and Jews most liberal (Oldmixon and Calfano Reference Oldmixon and Robert Calfano2007; McTague and Pearson-Merkowitz Reference McTague and Pearson-Merkowitz2013). However, prior research suggests that Catholics find abortion more morally objectionable than both mainline and Evangelical Protestants (Scott Reference Scott1989) and are more likely to strongly oppose the procedure (Scott and Schuman Reference Scott and Schuman1988). Moreover, to date, only the Catholic Church has spoken out specifically against IVF (Fortin and Abele Reference Fortin and Abele2016). A recent study on Senators’ voting patterns finds that Evangelical Protestant and Jewish office holders are the most constrained in their views on culture wars issues, including abortion, with Evangelical Protestants in opposition and Jews most supportive (McTague and Pearson-Merkowitz Reference McTague and Pearson-Merkowitz2015). Indeed, Jewish Americans have traditionally been more pro-choice (Cook, Jelen and Wilcox Reference Cook, Jelen and Wilcox1992; Jelen and Wilcox Reference Jelen and Wilcox2003), and Jewish organizations have advocated in support of ESC research (Holland, Lebacqz and Zoloth Reference Holland, Lebacqz and Zoloth2001), while internationally, Israel had the highest number of fertility clinics per capita and the highest rate of IVF treatments (Kahn Reference Kahn2000).

H3: Compared with mainline Protestants, Evangelical Protestants and Catholics will be more likely to describe all three procedures as morally wrong, while Jewish respondents will be more likely to describe them as morally acceptable.

H4: As with religiosity, the relationship between both religious denomination and attitudes about embryonic politics issues will be strongest for abortion and weakest for IVF, with ESC research in the middle.

Morality Politics and Attitude Constraints

To better understand the factors that shape attitudes about each of these procedures, as well as areas of consistency and inconsistency, I begin by examining individuals with constrained views about these embryonic policies. To do so, I conduct a multivariate analysis, dividing respondents into three categories: individuals who find all policies to be either morally wrong (0) or morally acceptable (2), compared with respondents with mixed views (1). Then, I examine each procedure separately to better understand the factors underlying these views, contrasting the results to see whether the same factors lead to opposition across policy areas.

Per H2 through H4, I expect that both religiosity and religious denomination will significantly influence an individual's views on the morality of each of these procedures, with the strongest relationship between these variables and attitudes about abortion, followed by ESC research, and then IVF. To capture religiosity, I include a series of dummy variables indicating how important religion is to the respondent, ranging from very important to not at all important (excluded as the reference group) (see the Appendix for a complete list of variables and coding). Another series of dummy variables measures religious denomination, including mainline Protestants (reference group), Evangelical Protestants, Catholics, and Jewish respondents, with the expectation that Evangelicals and Catholics will be more opposed to these procedures while Jews will be most accepting. I also include dummy variables for respondents who identify with the Church of Jesus Christ of Latter Day Saints (LDS/Mormons); as not religious; or with another religion, including Muslim, Buddhist, Hindu, Unitarian, and “other Christian.”Footnote ⁸

The multivariate models also include a number of control variables, including race/ethnicity, age, gender, education, and income. With respect to gender specifically, existing research largely finds no relationship with abortion attitudes (Cook, Jelen and Wilcox Reference Cook, Jelen and Wilcox1992; Jelen and Wilcox Reference Jelen and Wilcox2003) or ESC research (Ho, Brossard and Scheufele Reference Ho, Brossard and Scheufele2008, Shih, Scheufele and Brossard Reference Shih, Scheufele and Brossard2013). In contrast, the limited scholarship on IVF finds that in the United States, women are more likely to support insurance coverage for the procedure, while cross-nationally, they are more likely to think that associated costs should be reimbursed (Adashi et al. Reference Adashi, Cohen, Hamberger, Jones, De Kretser, Lunenfeld, Rosenwaks and Van Steirteghem2000; Nachtigall, MacDougall, Davis and Beyene Reference Nachtigall, MacDougall, Davis and Beyene2012). Similarly, I anticipate that women will be more likely to consider IVF morally acceptable. Given the cost barriers to IVF, and perceptions that it is primarily an upper-class phenomenon (Bell Reference Bell2009), I anticipate that socioeconomic status will play a greater role in IVF attitudes compared with other embryonic policy areas. In the models that follow, income is measured as a series of dummy variables (under $30,000 (reference group); $30,000–50,000; $50,000–75,000; $75,000–100,000; and over $100,000), while education is measured continuously. I include an additional control variable for marital status with the expectation that married respondents will be more likely to describe abortion as morally wrong and IVF as morally acceptable.

Because ideology and partisanship are not always consistent, I include dummy variables for both partisanship (Democrat, Republican, and independent, with independent as the reference group) and ideology (conservative, liberal, and moderate, with moderate as the reference group), with the expectation that respondents who are Republican and more conservative will be more likely to oppose these procedures, and particularly abortion, while liberals and Democrats will be more accepting.

Consistent Attitudes about Embryonic Policies

The first analysis seeks to understand the factors that are associated with constrained attitudes about embryonic policies, including respondents with the most consistent attitudes in either direction. Because morally acceptable, morally wrong, and not a moral issue cannot be clearly ordered, Table 1 presents the results of a multinomial logistic regression analysis comparing respondents who have mixed attitudes about the three procedures (baseline) with respondents who consider all three procedures to be either morally wrong (column 1) or morally acceptable (column 2). As multinomial logit coefficients are difficult to interpret, marginal effects for all statistically significant variables are included in the appendix (see Table B1).Footnote ⁹

Table 1.

Multinomial regression results: are all three embryonic policies morally unacceptable/morally acceptable (compared with mixed views on different policies)?

Robust standard errors are in parentheses. Significance levels: ^p < 0.10, *p < 0.05, **p < 0.01, ***p < 0.001.

Consistent with H2, individuals who consider religion to be very important are more likely to describe all three policies as morally wrong. However, there is no statistically significant relationship between religiosity and consistently describing the procedures as morally acceptable. Religiosity clearly shapes moral opposition to—but not support for—embryonic politics. Compared with mainline Protestants, Evangelical Protestants hold more constrained views, being nine percentage points more likely to describe all three procedures as morally wrong and five percentage points less likely to describe them as morally acceptable. Despite expectations that Catholic respondents would similarly be opposed to all three procedures, while Jewish respondents would be more accepting, these variables do not reach statistical significance in the first model. Interestingly, in comparison with mainline Protestants, LDS respondents are less likely to have constrained attitudes, being less likely to describe all three procedures as both morally acceptable and morally unacceptable. As discussed below, the LDS anomaly is likely the result of higher levels of support for IVF within this community. Yet, the small sample size of LDS (n  =  51) respondents suggests that these results should be interpreted with caution.

Consistent with existing research demonstrating that Latinos hold more conservative views on abortion and other family-related issues (Leal Reference Leal, Espino, Leal and Meier2007; Abrajano and Alvarez Reference Abrajano and Alvarez2010; Gibson and Hare Reference Gibson and Hare2012), Latino respondents are significantly more likely to oppose all three procedures. Demonstrating the importance of ideology, compared with moderates, conservative respondents are more likely to describe all three policies as morally wrong, while liberals are less likely to do so. Similarly, conservatives are less likely to describe all three policies as morally acceptable, while in comparison to independents, Democrats are more likely to do so, yet the substantive effects of these variables are weak in comparison to the religion variables. Other control variables also have significant effects. For instance, older respondents are more likely to describe all three procedures as morally acceptable, and respondents with higher incomes are less likely to describe them as morally wrong.

The above analysis finds support for H2 and parts of H3, demonstrating a relationship between religiosity and Evangelical Protestantism with constrained attitudes opposing embryonic policies. Yet, these results do not provide the opportunity to compare the extent to which the same factors influence attitudes about specific procedures. Next, I examine public opinion toward each procedure separately. Table 2 presents results of three multinomial logit models with attitudes about abortion, ESC research, and IVF included as distinct dependent variables. The models compare respondents who describe each policy as morally wrong (0) or morally acceptable (2) with a baseline of respondents who say it is not a moral issue (1). Given that I am most interested in understanding whether or not respondents consistently oppose these procedures, and the extent to which religion and religiosity influences this moral view, the discussion below will focus primarily on those results.

Table 2.

Multinomial regression, are abortion, ESC research, and IVF morally wrong or morally acceptable (compared with not a moral issue)

Robust standard errors are in parentheses. Significance levels: ^p < 0.1, *p < 0.05, **p < 0.01, ***p < 0.001.

Looking across the models, the strength of the religiosity variable is clear; across all policy areas, respondents who describe religion as a very important part of their lives are more likely to describe a procedure that involves the destruction of embryos as morally wrong. Figure 3 presents the marginal effects of describing oneself as very religious on attitudes toward the morality of each of the embryonic politics procedures. As the graphs in Figure 3 demonstrate, while there is a statistically significant relationship between religiosity and attitudes toward each of these procedures, as anticipated, the substantive effects are strongest with respect to abortion attitudes; compared with respondents who say religion is not at all important, those who say religion is very important are 30 percentage points more likely to characterize abortion as morally wrong, a dramatic shift. By contrast, respondents who are most religious are 16 percentage points more likely to categorize ESC research as morally wrong, and 10 percentage points more likely to do so for IVF. As Figure 3 demonstrates, the difference is most dramatic among the first two categories (religion being very important and somewhat important), with little difference in the latter categories (religion is not too important or not at all important).

Figure 3.

(Color online) Marginal effect of religiosity on public opinion that abortion, ESC research, and IVF are morally wrong.

H3 anticipates that, compared with mainline Protestants, Evangelical Protestants and Catholics will be more likely to describe these procedures as morally wrong, while Jews will find them morally acceptable, while H4 anticipates that the relationship between denomination and embryonic politics attitudes will be strongest with abortion and weakest with IVF. These results demonstrate partial support for this hypothesis. Evangelical Protestants are more likely to describe both abortion (17 percentage points) and ESC research (14 percentage points) as morally wrong, and again these results are both statistically and substantively significant. Even as Table 1 demonstrates more constrained attitudes across policy areas for Evangelical Protestants, surprisingly, the relationship between this variable and IVF attitudes is not statistically significant. In contrast to expectations that Catholics would be more likely to find these procedures unacceptable while Jewish respondents would find them more acceptable, in most cases, these variables are not statistically significant; the only exception is that Jewish respondents are more open to abortion. Because I am most interested in understanding which respondents describe these procedures as morally wrong, as a robustness check, I ran the models using logistic regression, where 1  =  morally wrong and 0  =  morally acceptable/not a moral issue. In this analysis, Catholics are more likely to oppose abortion while the variable for Jewish respondents becomes marginally significant (see Appendix Table B9). Surprisingly, LDS respondents are not more likely to oppose abortion or ESC research, but they are more likely to describe IVF as morally acceptable, and marginally less likely to describe IVF as morally unacceptable (p < 0.006). Although church guidelines oppose abortion in most cases, the church notes that with respect to IVF “(r)esponsibility for the decision rests solely upon” a husband and wife, meaning that this distinction is not driven by the church doctrine (The Church of Jesus Christ of Latter-Day Saints ND). The emphasis placed by the church on large families may contribute to more open attitudes about this procedure. Again, given the small sample size of LDS respondents, these results should be interpreted with caution, but this unexpected finding underscores that in addition to religion, ideas about the role of science in reproduction and other norms are also likely to influence attitudes about IVF.

Compared with the baseline category of not a moral issue, women are less likely to describe abortion as morally acceptable, but they are also less likely to describe it as morally wrong, though the latter relationship is only marginally significant. Notably, while Scott (Reference Scott1989) finds gendered differences in attitudes about the morality of abortion, these differences disappear in response to questions about abortion policy. There are no gendered differences in opinion about ESC research. Women are less likely to describe IVF as morally wrong and more likely to describe it as morally acceptable, though the latter result exceeds conventional measures of statistical significance. This result is particularly interesting in light of divisions within feminist groups about IVF, and ARTs more generally (Heidt-Forsythe Reference Heidt-Forsythe2013). While attitudes about gender roles are also thought to interact with religious beliefs (Cassese and Holman Reference Cassese and Holman2017), unfortunately, there is no variable included in the dataset to measure these views.

Although African Americans and Latinos have traditionally been more pro-life than white respondents (Combs and Welch Reference Combs and Welch1982; Uhlaner and Garcia Reference Uhlaner, Garcia, Norrander and Wilcox2002; Jelen and Wilcox Reference Jelen and Wilcox2003; Leal Reference Leal, Espino, Leal and Meier2007), some research suggests that the racial attitudes gap between Whites and African Americans may have shifted. Others find that variation in racial attitudes may be contingent upon religiosity (Strickler and Danigelis Reference Strickler and Danigelis2002; Carter, Carter and Dodge Reference Carter, Carter and Dodge2009) or gender (Combs and Welch Reference Combs and Welch1982; Wilcox Reference Wilcox1990). Across racial and ethnic groups, I find no difference in attitudes about abortion. Yet, African-American respondents are less likely to indicate that ESC research or IVF is morally acceptable, while Latinos are more likely to describe IVF as morally wrong.Footnote ¹⁰ Recall that in Table 1, Latinos are more likely to consistently oppose all three procedures. These results suggest that greater opposition to IVF among Latinos drive this high level of constraint.Footnote ¹¹ Compared with white respondents, individuals identifying with another race are also more likely to describe IVF as morally unacceptable. Unfortunately, little information is known about the background of individuals in this category, but these findings point to the need for further research on variation in attitudes about a range of embryonic policies by race/ethnicity.

Despite my expectations of a stronger relationship between income and IVF, as compared with abortion or ESC research, I find that the opposite is the case. In general, individuals in the highest income category are less likely to view all three procedures as morally wrong, but the relationship is strongest (−0.18) with respect to abortion and weakest (−0.09) for IVF.

The results in Table 2 also reiterate the important role of ideology in morality politics. Respondents who describe themselves as conservative are significantly more likely to describe each of the three procedures as morally wrong, and liberals are less likely to do so, though the results for ESC research are only marginally significant. Existing research demonstrates strong partisan polarization surrounding abortion and stem cell research (Adams Reference Adams1997; Jelen and Wilcox Reference Jelen and Wilcox2003; Hillygus and Shields Reference Hillygus and Shields2014). I find that Democrats are more likely to describe all three procedures as morally acceptable and are less likely to describe abortion as morally wrong. Surprisingly, the dummy variable for Republicans is significant only for ESC research. Consistent with higher rates of abortion among unmarried women, married respondents are also more likely to describe abortion as morally unacceptable. In contrast, the absence of a relationship between marital status and attitudes about IVF is unexpected.

Given the relatively low levels of debate surrounding IVF, it is possible that some respondents indicating IVF is not a moral issue are unfamiliar with the specifics of the procedure and that embryos may be destroyed in the process. Unfortunately, in the present survey, no follow-up questions were asked that would provide the opportunity to gauge respondents’ knowledge about IVF. Notably, the sample size is largest with respect to the IVF question, suggesting that at a minimum, people's lack of familiarity did not prevent them from answering questions about the procedure. In light of the lack of research surrounding public opinion on IVF, my findings represent an important step in improving our understanding of attitudes about this procedure. One important avenue for future research would be to better understand people's underlying knowledge about IVF and the way that this may influence attitudes about its morality.